Each year approximately 2 million people suffer traumatic brain injuries (TBI) in the U.S. Of these about 100,000 die and 90,000 are left with long-term disabilities. Advances in the management of these patients have reduced mortality but done little to ameliorate brain injury. Recent studies in severe TBI patients have suggested that reduced regional cerebral blood flow (rCBF) in the first few hours after injury contributes to secondary brain injury. Additionally, the use of acute hyperventilation (HV) to treat elevated intracranial pressure following TBI may led to or exacerbate ischemia thus augmenting rather than preventing secondary CNS injury. In order to understand the impact of HV on brain oxygenation, it is important to determine if the reduction in rCBF seen in TBI patients produces brain oxygen deprivation severe enough to cause ischemic neuronal damage. The presence of a PET scanner in the NNICU at Barnes-Jewish Hospital, combined with our extensive experience with the use of PET to detect ischemia, our expertise in the clinical management of TBI patients, and the large available patient population gives us a unique opportunity to address these issues. We propose to investigate the impact of acute HV on cerebral blood flow and brain oxygenation in TBI. Specifically we will test the hypothesis that severe (PaCO2 of about 25 mm Hg) but not moderate (PaCO2 of about 30 mm Hg) hyperventilation following TBI can produce reduction in CBF potentially severe enough to cause cerebral energy failure (defined as CBF that is insufficient to meet metabolic needs). We will study 30 patients within 12 hours of non-penetrating severe head injury (GCS less then or equal of 8) who have just completed PET measurements of regional (CBF, CMRO2, OEF2 A-VDO2 and CvO3 as part of project 1. Patients with elevated ICP(18 - 25 mm Hg) at the time of the initial PET study will have repeat PET measurements of rCBF and CMRO2, OEF, A-VDO2, and CvO2, again after acute HV, Three groups of 10 patients each will be studied. The first group will have PaCPO2 lowered to 30+2 torr. If no patient develops cerebral ischemia potentially severe enough to cause energy failure the second group will have PaCO2 lowered to 25+2 torr. Otherwise, the second group will have PaCO2 lowered, in those patients with an initial PaCO2 greater then or equal to 35 torr, to 35=2 torr to determine if there is a safe threshold for HV. A third group of ten patients will serve as non- hyprventilated controls. These investigations are critical to determine the proper use of HV in the treatment of acute TBI.